Statistical Forecasting of Fresh Vegetable Sales Trends: Application of Correlation Analysis and ARIMA Model

Abstract

Vegetable commodities have many varieties, short shelf life special purchase times, and other status quo, so the day not sold out of vegetables often can not be sold the next day. Therefore, it is important to accurately predict the demand for vegetable commodities in the coming period and the amount of demand for the replenishment and profitability of vegetable superstores. In this study, Spearman's correlation analysis was applied to investigate the relationship between the daily sales of six major vegetable categories: leafy vegetables, cauliflower vegetables, aquatic root vegetables, eggplant vegetables, pepper vegetables, and edible mushrooms, and the results showed that there was a significant positive correlation between aquatic roots and edible mushrooms which The results show that there is a significant positive correlation between aquatic root vegetables and edible mushrooms, which indicates that consumers who buy aquatic root vegetables tend to buy edible mushrooms as well and that vegetable superstores should take note of this potential relationship to better formulate replenishment strategies; and then the ARIMA model was applied to predict the demand for these six major vegetable categories in the coming week, and the average relative error of the test set was 24.93%, which is a good prediction effect.